A tailor-made development for time domain data series pre-processing in the power industry

Abstract

This paper introduces a tailor-made development, supported in Matlab, to handle a set of instrumental observations associated to energy production technologies. Original sequence of values can be obtained by a filtered one and/or by its curve of trends. Accuracy knowledge of its time scale signal is not strictly necessary because instead of applying signal processing theory and methods to a window of data values, the statement is translated into statistic methods applied to a data time series. A Graphical User Interface Development Environment (GUIDE) integrates data loading, filters configuration, visualization facilities to validate filtering and saving data, resulting in a powerful pre-processing tool. This software platform solution will facilitate the application of such techniques as modelling and identification, monitoring chemical processes, building tools for control strategy decisions, and failure diagnosing, among others.